Vehicle Compartment Low Frequency Coupling Noise Prediction Analysis And Acoustic Optimization Based On Approximation Modal

Vehicle acoustics comfort has become an important indicator to measure vehicle quality.Good acoustic environment of vehicle compartment is the necessary condition to ensure vehicle ride comfort.For the vehicle compartment low frequency noise problems which are both prominent and difficult to eliminate,the main research works of this paper are as follows:1.The finite element modal of body-in-white,a structure FEM for vehicle body and an acoustic FEM for in-car cavity were established.And then,the modal analysis of white body and vehicle structure,the acoustic modal analysis for the cavity sound field and the coupling modal analysis for the structure-acoustic coupling system were analyzed respectively.Through the analysis to know about the modal characteristic of the vehicle body and the sound field,the analysis results showed that there are many local modes in the areas of the door of right and left,front rail panel,the roof and the floor.In the frequency range of 157Hz-161 Hz,structure-acoustic coupling system modes are distributed intensively and easy to cause resonance in the action of the external excitation.2.The time domain model of road random excitation was established by using white noise filtering method to simulate the road random excitation,and seven degree of freedom vibration dynamic modal for vehicle was set up.Matlab/Simulink tools were utilized to calculate the suspension excitation force,then by using fast Fourier transform method the suspension excitation force amplitude frequency spectrum was obtained.For the structure-acoustic coupling system,the acoustic coupling vibration response which based on modal analysis was carried out under the action of the excitation force of engine and suspension to predict passenger compartment acoustic environment.The analysis results indicate that the measuring point has high sound pressure at the frequency of 158 Hz,134Hz,198 Hz,152Hz,78 Hz.3.According to the principle of the panel acoustic contribution analysis,vehicle body panel was divided into 24 pieces.Taking panel's vibration velocity as boundary conditions,panel acoustic contribution analysis was carried out.In the frequency of high sound pressure,the panels that contribute largely to the measuring point sound pressure was identified as follows: 2_dp_q,7_dpw,8_qw,9_db_q,10_db_z,11_db_h,12_cm_zq,13_cm_zh,14_cm_yq,15_cm_yh.They namely are the floor,the roof,thefront rail panel and the inner door panel.4.Taking the biggest contribution panel's thickness parameter as factor,vehicle weight,the seventh order modal frequency of vehicle body,measuring point peak sound pressure and sound pressure root-mean-square value as response,the experiment design was performed by using Optimal Latin Hypercube Design method in the factor design space,and a total of 30 groups of sample data were collected.Approximation modal was established by using the RBF neuro-network method,and the error of the approximation modal was analyzed.The results showed that the RBF neuro-network modal that was builted with small errors and high accuracy can meet the requirements of modeling,and can also replace the finite element model to perform the optimization.5.Taking panel's thickness parameter as design variable,the minimize of the peak sound pressure of measuring point as optimal object,sound pressure root-mean-square value,vehicle weight and the seventh order modal frequency of vehicle body as constraint condition,the panel thickness optimization design was carried out by using the RBF neuro-network model and the adaptive simulated annealing algorithm.Optimization results showed that the measuring point D's peak sound pressure reduced by 4.45 dB at frequency 158 Hz and decreased by 5.47 dB at 134 Hz.In other frequency,sound pressures of measuring point were obviously decreased too.It indicated that through optimizing the car chamber pressure reduces effectively,and the car room acoustical comfort is significantly improved.It also proved that through the establishment of RBF neuro-network approximation model,the design method that using the adaptive simulated annealing algorithm to optimize the vehicle plate thickness parameter is feasible and effective.